garage
garage copied to clipboard
A toolkit for reproducible reinforcement learning research.
Is there an easy way to get specific data from MetaEvaluator? For example, pulling out the return for each episode in each meta-testing task? This would allow looking at a...
I'm sorry to bother you, but I encountered an error, ``` File "/mnt/data/chenjinhao/.local/lib/python3.7/site-packages/tensorflow/python/client/session.py", line 1161, in _run np_val = np.asarray(subfeed_val, dtype=subfeed_dtype) ValueError: setting an array element with a sequence. ```...
Hi. I'm currently developing some other algorithm using TRPO. I found that the training keeps failing after certain epochs because of CUDA out of memory error. At first I thought...
The following line of code was present twice self.replay_buffer = replay_buffer
Hi Garage Developers, The DQN algo uses 1 worker for running _num_eval_episodes, which seems to be a bottleneck in its performance. Has anyone tried leveraging all the compute cores to...
Hey all, I am having some trouble reproducing the baseline numbers on both Atari and Fetch (Push/Pick-and-Place/Slide) tasks. Although, the same code works for Pong, Enduro, it seems to not...
Hi all, I've used the deterministic.set_seed but still get different results. It seems that the samples that start from itr2 become different (itr0 & itr1 are the same). Do you...